i'm pretty new to boost and ublas but i'm trying to integrate the boost::numeric::ublas::vector into a group simulation code.

Mainly we are concerned with cartesian vectors so i have designed a class, BoostVector, that uses a vector<gfloat, boost::numeric::ublas::bounded_array<double, 3> > for storage/manipulation. I'm also using the operators.hpp to generate operations between BoostVector objects that call operations defined by the ublas headers. 

the error i get is when i attempt to perform a subtraction between two BoostVector objects. One vector has size 3, as expected, the other has size 86. I dont' understand how the second vector can even be resized to 86 given that i'm using bounded_array..... this vector is involved in matrix multiplication with 3x3 matrices...... so i'm at a loss to explain the size... 

i should note that right before the error the value of the size 86 vector may be anomalous, i.e. on the order of e-386..... this may represent an error or just machine precision issues...

Tony